Predicting dropout in dietary weight loss trials using demographic and early weight change characteristics: Implications for trial design

摘要: Summary Attrition causes analytical and efficacy issues in weight loss trials. Consistent predictors of attrition trials have not been identified. Trial design could be improved if factors predicting are accounted for. The aim this study is to quantify the effect easily measured pre early variables determine their relationship with dietary Methods Data was pooled from four previous Mixed effects logistic regression, Receiver Operator Curves decision trees (classification regression trees) were used which (percent at 1 month, age, gender baseline BMI) predicted dropout cutoffs useful for future trial design. Results sample included 289 subjects, 73% female, a mean age 46.68±9.27years average 25%. Percent month strongest predictor dropout, those ≤2% 4.99 times (95% CI 2.71, 9.18) more likely drop out than >2% first ( P 50 =0.006). Discussion Early identified as significant designs maybe by incorporating these developing interventions targeting may improve participant retention.